Feeding practice may influence outcomes of major clinical significance for preterm infants in neonatal intensive care units. These outcomes include dietary intake and somatic growth. Although the scientific basis of the nutrition for preterm infants is well established, little is known about how feeding practice for the newly nipple-feeding preterm infant influences these feeding outcomes. The objective of the proposed study is to examine the effect of two components of feeding practice--regulation (i.e., prescribed vs. ad libitum feeding) and dietary type (i.e., 20-calorie/ounce formula vs. 24-calorie/ounce formula) on dietary intake of preterm infants who are mature and medically stable enough to be completely nipple fed. The propose study will use a 2 X 2 experimental design with random assignment of infants to either ad libitum or prescribed feeding practice and, within the feeding practice, to 20- or 24-calorie/ounce formula. At least 18 infants will be randomly assigned to each of the four study groups. All infants admitted to the NICU born at less than 35 weeks gestation with weight that is appropriate for gestational age and who are exclusively bottle fed will be eligible for this study. Study infants must not have congenital or medical problems interfering with nutritional status. Infants may have recovered from Respiratory Distress Syndrome (RDS) or have mild Bronchopulmonary Dysplasia (BPD). Infant characteristics that will be controlled statistically are: (1) pulmonary health status; and (2) postconceptional age and on the second day of full nipple feeding. The infant's 24-hour dietary intake and daily weight will be the basis for the following measures: (1) volume in cc/kg/day of intake; (2) kcalories/kg/day of intake; and (3) g protein/kg/day of intake. Two variables will be used to assess somatic growth: (1) grams weight gain/day divided by the mean weight for the 7-day study period; and (2) mm/day growth in mid-arm muscle circumference. Study hypotheses will be tested with MANOVA with five dependent measures. The interaction of the two feeding practice factors (feeding regimen and caloric density) will be examined.